REGULATION OF NUCLEOCYTOPLASMIC SHUTTLING OF THE TRANSCRIPTIONAL REGULATOR HUMAN MESODERM INDUCTION EARLY RESPONSE 1 α IN BREAST CARCINOMA CELLS

(1)

REGULATION OF NUCLEOCYTOPLASMIC SHUTTLING OF THE TRANSCRIPTIONAL REGULATOR HUMAN MESODERM INDUCTION

EARLY RESPONSE 1 α IN BREAST CARCINOMA CELLS

By

© Shengnan Li A Thesis submitted to the School of Graduate Studies

In partial fulfillment of the requirements for the degree of Doctor of Philosophy

Division of Biomedical Sciences Faculty of Medicine

Memorial University of Newfoundland

(May 2018)

St. John’s Newfoundland and Labrador

(2)

i

Abstract

Temporal and spatial regulation of the subcellular distribution of

transcriptional regulators is important to ensure their proper functioning in a cell.

Mesoderm induction early response 1 α (MIER1α) has been implicated as a tumour suppressor in breast cancer. Analysis of MIER1α subcellular localization in breast samples revealed a stepwise translocation from the nucleus to the cytoplasm during progression to invasive carcinoma (McCarthy et al., 2008). Therefore, an

investigation of MIER1α nucleocytoplasmic shuttling is critical to unraveling its role in breast cancer progression.

Structurally, MIER1α has conserved domains found in a number of other transcriptional regulators, including N-terminal acidic stretches, ELM2 and SANT domains. However, none of these domains contain the predicted nuclear import or export signals. In this thesis, I show that MIER1α localizes in the nucleus in breast carcinoma MCF7 cells without an intrinsic nuclear localization signal (NLS).

Although MIER1α has been shown to bind to ERα, active nuclear import of MIER1α is not through interaction with ERα; instead, it depends on interaction and co- transport with HDAC1/2 through a “piggyback” mechanism. Deletion analysis demonstrated that the entire ELM2 (aa164-283) is required and sufficient for nuclear targetting of MIER1α and that a simple mutation, ²¹⁴W→A in the ELM2 domain abolishes both the interaction between MIER1α and HDAC1/2 and its nuclear localization.

(3)

ii

Further investigation revealed that MIER1α is exported out of the nucleus when cells are treated with insulin, IGF-1, EGF or FGF, but not with 17β-estradiol, and this export out of the nucleus is mediated by CRM1. HDAC1 & 2 nuclear

localization were not affected by MIER1α export, suggesting they are only involved in MIER1α nuclear import. Both Mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinase B/Akt (PI3’K/AKT) pathways are activated upon

treatment with growth factors, and it was further confirmed MIER1α nuclear export is triggered by the MAPK pathway, but not the PI3’K/AKT pathway. However, the mutation of predicted ERK1/2 consensus phosphorylation sites S¹⁰-P and/or S³⁷⁷-P motifs in the MIER1α sequence had no effect on its localization. MIER1α returns to the nucleus when activation of MAPK pathway diminishes, suggesting this process is transient and reversible. Deletion analysis narrowed the required sequence for export to the N-terminal region, aa1-163, containing acidic stretches. Overall, these results provide details of the mechanism responsible for MIER1α nucleocytoplasmic shuttling in a breast cancer carcinoma cell line; a similar mechanism may be

operating during breast cancer progression.

(4)

III

Acknowledgements

I would like to express my deepest gratitude and respect to my supervisor, Dr.

Laura L. Gillespie for her supervision, support, and encouragement throughout the entire study. I am also very grateful to Dr. Gary Paterno for his constant guidance and help. Additionally, I would like to express my thanks to my committee members Dr. Ann Dorward and Dr. Ken Kao for their expertise and invaluable comments.

I would like to extend my thanks to the present and past lab members, Corinne, Roya, Amy, Leena, Julia, Satoko, Phil, Paula, and Youlian for the friendship and for providing a great working environment. Also sincere thanks to the Faculty of Medicine and the School of Graduate Studies for their cordial administrative help and support through this program.

My love and appreciation go to my parents, who instilled strong beliefs in me to pursue my dreams and took care of my daughter while I stayed abroad. My

appreciation is also due to my dearest husband and sisters, who encouraged me and supported me throughout this journey. My apologies go to my daughter for my absence during her infancy.

(5)

IV

List of Tables

Table 1.1 Highlights of genomic, clinical and proteomic features of breast tumour

subtypes ... 12

Table 2.1 List of cell lines used in this study ... 66

Table 2.2 Antibodies used for immunofluorescence (IF) and western blot (WB) ... 68

Table 2.3 ELM2 domain does not contain cNLS ... 85

Table 4.1 Primer sequences for site-directed mutagenesis of SP motifs on MIER1α ... 161

Table 4.2 SP motifs within the MIER1α sequence... 187

Table 4.3 MIER1 post-transcriptional modification reported in research papers ... 187

(10)

IX

List of Figures

Figure 1.1 Incidence of cancers in 2017 in Canada. ... 4

Figure 1.2 Schematic representation of the NPC ... 27

Figure 1.3 Ran directs nucleocytoplasmic transport ... 34

Figure 1.4 Structure of the human MIER1 gene and splice variants ... 40

Figure 1.5 Protein domains and motifs in hMIER1 isoforms... 48

Figure 1.6 Loss of nuclear MIER1α during breast cancer progression ... 53

Figure 2.1 MIER1α amino acid sequence ... 59

Figure 2.2 Amino Acid sequences of all plasmids used ... 61

Figure 2.3 Knockdown of ERα does not affect nuclear localization of MIER1α in MCF7 cells ... 73

Figure 2.4 MIER1α is localized in the nucleus in ER- breast carcinoma cells ... 76

Figure 2.5 The ELM2 domain directs nuclear localization of MIER1α ... 80

Figure 2.6 The ELM2 domain is sufficient for nuclear localization of MIER1α... 82

Figure 2.7 Nuclear localization requires an intact ELM2 domain ... 86

Figure 2.8 Statistical data demonstrate the necessity of intact ELM2 domain for MIER1α nuclear localization ... 88

Figure 2.9 Western blot showing that ELM2 mutant (²¹⁴W→A) does not interact with HDAC1 or HDAC2 ... 92

Figure 2.10 Interaction with HDAC1/2 is required for nuclear localization of MIER1α ... 94

(11)

X

Figure 2.11 ²¹⁴W→A mutation abolishes MIER1α nuclear localization ... 96

Figure 2.12 Individual and double knockdown of HDAC1 and 2 in MCF7 cells using shRNA ... 100

Figure 2.13 HDAC1 and 2 knockdown reduces nuclear localization of MIER1α ... 102

Figure 3.1 Insulin treatment reduces nuclear localization of MIER1α ... 115

Figure 3.2 IGF-1 treatment reduces nuclear localization of MIER1α ... 118

Figure 3.3 EGF treatment reduces nuclear localization of MIER1α ... 121

Figure 3.4 FGF treatment reduces nuclear localization of MIER1α ... 124

Figure 3.5 E2 has no effect on subcellular localization of MIER1α ... 127

Figure 3.6 HDAC1 and 2 localization are not affected by insulin ... 130

Figure 3.7 HDAC1 and 2 localization are not affected by IGF-1 ... 132

Figure 3.8 HDAC1 and 2 localization are not affected by EGF ... 134

Figure 3.9 HDAC1 and 2 localization are not affected by FGF ... 136

Figure 3.10 HDAC1 and 2 localization are not affected by E2 ... 138

Figure 3.11 LMB abolishes MIER1α nuclear export caused by insulin... 143

Figure 3.12 LMB rescues MIER1α nuclear loss caused by insulin and growth factors ... 145

Figure 4.1 The mammalian MAPK cascade ... 154

Figure 4.2 Schematic diagram depicting the most representative signaling of the PI3K/AKT pathway ... 156

Figure 4.3 MIER1α amino acid sequence and plasmids used in the study ... 159

(12)

XI

Figure 4.4 AKT VIII inhibits growth factor-dependent AKT activation ... 169 Figure 4.5 U0126 inhibits growth factor-dependent MAPK activation ... 172 Figure 4.6 AKT inhibition has no effect on MIER1α subcellular distribution... 175 Figure 4.7 U0126 inhibits growth factor dependent nuclear loss of MIER1α in MCF7 cells ... 178 Figure 4.8 ERK1/2 is not activated by E2 ... 182 Figure 4.9 Substitution of S¹⁰ to A has no effect on the subcellular distribution of MIER1α in MCF7 cells ... 188 Figure 4.10 Substitution of S³⁷⁷ to A has no effect on the subcellular distribution of MIER1α in MCF7 cells ... 190 Figure 4.11 Double mutations of ¹⁰S and ³⁷⁷S to A has no effect on the subcellular distribution of MIER1α in MCF7 cells ... 193 Figure 4.12 The N-terminus containing 4 acidic stretches is required for MIER1α nuclear export ... 197 Figure 4.13 The N-terminus containing 4 acidic stretches is required for MIER1α nuclear export ... 199 Figure 4.14 MIER1α nuclear loss caused by ERK1/2 activation is transient and reversible ... 203

Figure 5.1 Model of MIER1α nucleocytoplasmic shuttling mechanism ... 210

(13)

XII

List of Abbreviations and Symbols

α alpha

β beta

aa amino acid

Ada2 adaptor 2

AF-2 activation function-2

AID activation-induced deaminase

ALK Anaplastic Lymphoma Kinase

amp ampicillin

ANOVA analysis of variance

ARM armadillo

ATCC American Tissue Culture Collection

BAHD1 Bromo-Adjacent-Homology domain-containing 1

bp base pair

BRCA1 breast cancer susceptibility gene 1

BSA bovine serum albumin

CAT chloramphenicol acetyltransferase

CBP CREB-binding protein

CCL2 chemokine (C-C) motif ligand 2 CDKN1A cyclin-dependent kinase inhibitor 1A

CDKN1B CDK inhibitor 1B

CDKs cyclin-dependent protein kinases

CDYL chromodomain-containing protein

co-IP co-immunoprecipitation

co-REST REST co-repressor

CREB cAMP response element binding protein

CRM1 Chromosomal Maintenance 1

DAD deacetylase activating domain

oC degrees Celsius

DBD DNA binding helix-turn-helix domain

DCIS ductal carcinoma in situ

DMEM Dulbecco s Modified Eagle s Medium

DMSO dimethyl sulfoxide

DNA deoxyribonucleic acid

E2 17-β-estradiol

E. coli Escherichia coli

EGF epidermal growth factor

EGFR epidermal growth factor receptor

EHMT2 Euchromatic histone-lysine N-methyltransferase 2 Elf-1 ETS transcription factor

ELM2 EGL-27 and MTA1 homology domain 2

(14)

XIII

EMT epithelial-mesenchymal transition

ER estrogen receptor

ERα estrogen receptor α

ERK1 extracellular signal-regulated kinase-1 ERK2 extracellular signal-regulated kinase-2 ETS E-twenty six family transcription factor

FGFs fibroblast growth factors

FGFR fibroblast growth factor receptor FG-repeats phenylalanine-glycine repeats

G9a histone methyltransferase

GCN5 general control of amino-acid synthesis GSK-3 glycogen synthase kinase-3

h hour

HATs histone acetyltransferases

HDACs histone deacetylases

HEK human embryonic kidney

HER2 human epidermal growth factor receptor 2

HER2E human epidermal growth factor receptor 2 enriched

HGF hepatocyte growth factor

HGFR hepatocyte Growth Factor Receptor

HOX Homeobox

hTERT human telomerase reverse transcriptase

IBB importin-β binding

IDC invasive ductal carcinoma

IEGs Immediate early response genes

IHC immunohistochemistry

IGF Insulin-like growth factor

IMTs inflammatory myofibroblastic tumours

IP immunoprecipitation

JNKs Jun N-terminal protein kinases

kDa kilodalton

KLK3 prostate-specific antigen Kallikrein 3 LEF-1 lymphocyte enhancer factor 1

LMB Leptomycin B

LPL Lipoprotein Lipase

LXXLL Leu-Xaa-Xaa-Leu-Leu

M molar

MAPK mitogen-activated protein kinase

MC2 MDA-MB-231 derived ERα+ stable clone cell line MEK1 mitogen-activated protein kinase kinase 1

MEK2 mitogen-activated protein kinase kinase 2 MIER1 mesoderm induction early response 1 (gene) MIER1 mesoderm induction early response 1 (protein)

(15)

XIV

MIER2 mesoderm induction early response 2 (gene) MIER3 mesoderm induction early response 3 (gene)

min minute

miRNA microRNA

ml millitres

MTA-1 metastasis-associated protein mta1 metastasis-associated gene 1

mRNA messenger RNA

N-CoR nuclear receptor co-repressor

NES nuclear export signal

ng nanogram

NLS Nuclear localization signal

nM nanomole

NGF nerve growth factor

NPC nuclear pore complex

NTF nuclear transport factor 2

NTRK1 Neurotrophic tyrosine receptor kinase 1

Nups nucleoporins

NuRD nucleosome remodeling deacetylase PAGE polyacrylamide gel electrophoresis

PAS polyadenylation signal

PBS phosphate buffered saline

PDGF Platelet derived growth factor

PDK1 phosphoinositide-dependent kinase 1

PH pleckstrin-homology

PI3K phosphoinositide 3-kinase B

PIP3 phosphatidylinositol (3,4,5)-trisphosphate

PKB protein kinase B

PP2A protein phosphatase 2A

PPAR-γ peroxisome proliferator-activated receptor-γ

PR progesterone receptor

PTGS2 Prostaglandin-endoperoxide synthase 2 PTMs post-translational modifications

RanBP1 Ran-binding protein-1

RB retinoblastoma

RCC1 chromosome condensation

RNPs ribonucleoproteins

RTK receptor tyrosine kinase

rRNA ribosomal RNA

SANT Swi13, Ada2, N-CoR, TFIIIB

SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis

SH3 SRC Homology 3

SMRT silencing mediator of retinoic acid

(16)

XV

SPP1 secreted phosphoprotein 1

STAT1 signal transducer and activator of transcription 1

S/T-P Ser/Thr-Pro motif

STATs Signal Transducers and Activators of Transcription Swi3 switching-defective protein 3

T thymidine

T-ALL T-cell acute lymphoblastic leukemia

TFs transcription factors

TGF β transforming growth factor β

TMAS tissues microarrays

TNBC Triple-negative breast cancers

TPR translocated promoter region

TREX transcriptional export

tRNA transfer RNA

TSP tumour suppressor proteins

µg micrograms

µl microliters

VEGF vascular endothelial growth factor

VC5 MDA-MB-231 derived ERα- stable clone cell line

W tryptophan

xmierl xenopus mesoderm induction early-response 1 (gene) xmier1 xenopus mesoderm induction early-response 1 (protein)

(17)

XVI

List of Appendices

Appendix 1: EndoFree^® Plasmid Maxi Kit Protocol ... 242 Appendix 2: Immunofluorescence ... 244 Appendix 3: In silico analysis of MIER1α amino acid sequence by NetNES 1.1 ... 245

(18)

1

Chapter 1 General Introduction

1.1 Cancer overview

In mammals and other multi-cellular organisms, the activities of normal cells are tightly regulated by signals in their surroundings. The signals released through paracrine, autocrine or endocrine mechanisms ultimately dictate a cell’s fate: to grow, to differentiate, to proliferate or to undergo apoptosis. The plethora of signalling networks and interconnecting factors ensure the functional homeostasis of our bodies. A healthy cell is in harmony with its environment by responding and integrating the external messages through a highly regulated signal transduction network. This balance, however, can be destroyed and can result in the development of cancer.

Cancer cells are characterized by uncontrolled proliferation (Hanahan &

Weinberg, 2011). Rather than responding to the signals that control normal cell behaviour, cancer cells proliferate in an uncontrolled manner, eventually resulting in accumulated abnormalities in several aspects of cell behaviour which distinguish cancer cells from normal cells.

1.1.1 Types of cancer

Types of cancer are usually named for the organs or tissues from which the cancers arises; they may be also further described by the subtypes of cancer

initiating cells, and how they are characterized histologically. But in brief, one of the

(19)

2

most important issues in cancer pathology is the distinction between benign and malignant tumours. A benign tumour remains in its original location and does not invade the surrounding normal tissue or spread to distant body sites, while

malignant tumours can invade normal tissue and spread through the body via blood circulatory or lymphatic systems, in a process called metastasis. Loss of primary organ function or distant organ function as a result of cancer metastasis threatens physiological homeostasis, leading to death of the patient.

Malignant tumours can arise in virtually any part of the body but fall into one of six main groups, including: carcinomas, sarcomas, myeloma, lymphomas,

leukemia and mixed types. The most common cancers are carcinomas, accounting for 80 to 90 % of all cancer cases (Canadian Cancer Society, 2017). Carcinomas originate from epithelial cells, which either cover surfaces or line internal organs such as skin, breast, prostate or lung. Sarcomas are solid tumours appearing in connective tissues, such as muscle, bone or fibrous tissue; while only accounting for less than 1% of all adult solid malignant cancers (Burningham, Hashibe, Spector, &

Schiffman, 2012), they are often fatal. Unlike solid tumours, leukemias are cancers arising from the blood-forming cells or immune system cells. They manifest as the overproduction of white blood cells and account for approximately 8% of human malignancies (Canadian Cancer Society, 2015).

Cancer is an age-related disease (Fig. 1.1A) and 70% of new cancer cases will occur in Canadians aged 50 years or older (Statistics Canada, 2014). Cancers

occurring in ten different body sites account for more than 75% of total cancer

(20)

3

incidence; among these breast, prostate, lung and colon/rectum are the four most common cancer sites, accounting for more than half of all cancer cases in men and women (Fig. 1.1B). Fig. 1.1 is the cancer incidence in Canada reported by Canadian Cancer Society (Canadian Cancer Society, 2017).

(21)

4 Figure 1.1 Incidence of cancers in 2017 in Canada.

(Modified from (Canadian Cancer Society, 2017) with permission) (A) Distribution of 10-year tumour-based prevalence for selected cancers, Canada, January, 1, 2009;

(B) Distribution of new cancer cases for selected cancers by age group, Canada, 2009–2013.

(22)

5

(23)

6 1.1.2 The development of cancer

The molecular view of cancer today is that cancer develops over time rather than all at once, as a result of the cumulative effect of genetic changes (Podlaha, Riester, De, & Michor, 2012). Each change causes the cells to acquire some traits and the accumulation of these changes altogether promotes the malignant growth of cancer cells. There is clear evidence that the control of cell cycle, cell survival and the elimination of unnecessary or damaged cells in normal cellular programs are altered during tumourigenesis.

A single point mutation is not sufficient to generate a cancer cell from a preexisting normal cell. Two main types of genes play a major role in triggering normal mouse cells to be transformed: proto-oncogenes and tumour-suppressor genes (Alberts, 2008). Proto-oncogenes usually regulate cellular growth. On the other hand, tumour-suppressor genes inhibit cell division, promote apoptosis, or both. Proto-oncogenes and tumour-suppressor genes coordinate to regulate the growth of each tissue and organ in the body. In rat models, the disturbance of cell growth control and cell morphology can transform normal rat cells into cancerous cells (Alberts, 2008). For example, the co-introduction of myc, a gene that helps cells to become immortalized, and ras oncogene, which changes the morphology of cells to rat embryo fibroblasts, yields a foci of transformants (Wang, Lisanti, & Liao, 2011).

The separate existence of myc or ras, would not result in these transformants in human cells (Wang et al., 2011); transformation of human cells usually requires the collaboration of more than one mutated gene. Experimental results imply that five

(24)

7

distinct cellular regulatory circuits need to be altered for normal human cells to develop into tumour cells (Alberts, 2008). The first change is the the induction of the human telomerase reverse transcriptase (hTERT) gene, as this gene is necessary to maintain the telomere. The other four changes involve: (1) the mitogenic signalling pathway controlled by Ras-like genes; (2) the cell cycle checkpoint controlled by pRb; (3) the guarding pathway controlled by p53; and (4) the signalling controlled by protein phosphatase 2A (PP2A). However, the necessity of these five changes was determined by in vitro experiment; it is still unclear whether the steps needed in vitro reflects the changes that occur in vivo and lead to cancer in humans.

1.1.3 Hallmarks of cancer

The traits that cancerous cells acquire are called the “Hallmarks of Cancer.”

These traits are the characteristics that distinguish cancer cells from normal cells.

When cell division and the cell death are both interrupted by external factors, normal cells can then overwhelm the body’s defenses and become cancerous. The hallmarks of cancer are described in two landmark scientific papers by Douglas Hanahan of the University of California and Robert Weinberg of the Massachusetts Institute of Technology (Hanahan & Weinberg, 2000, 2011).

1. Self-sufficient cell division The most fundamental trait of cancer cells is that they are able to sustain proliferation. Growth-promoting signals in normal tissues are carefully controlled and released to maintain a homeostasis of cell number and thus normal tissue architecture. Cancer cells, on the contrary, control their own

(25)

8

proliferation by producing growth signals themselves or by elevating signal receptors.

2. Resistance to anti-growth signals Cell proliferation is stimulated by signals; also, some signals “put the brakes” on cell growth and proliferation. Once growth

inhibition is interrupted or ignored, cancer cells can have unlimited proliferation.

Unlimited proliferation is usually a result of mutations or alterations of tumour suppressor genes and proto-oncogenes.

3. Evading programmed cell death Cells are programmed to die in the event they become damaged, a mechanism called apoptosis to prevent the propagation of DNA errors. On the other hand, the apoptosis signals of cancer cells can be disrupted when tumour suppressor genes suffer mutations or other damage.

4. Limitless replicative ability A solid tumour may be composed of billions of cells, an indication of uncontrolled cell division. The telomere is a small portion located at the end of each chromosome. In normal cell division, the telomere is shortened every time DNA is replicated because the end of the telomere cannot be covered by the Okazaki fragment and therefore, get copied. Ultimately, telomeres reach a critical point and the cell can no longer divide. Telomerase is an enzyme which can maintain telomere length and when it is activated in a cancer cell, telomerase will allow it to replicate indefinitely compared to normal non-cancerous cells with self-limited replication.

(26)

9

5. Sustained angiogenesis The development of new blood vessels is called

angiogenesis. Angiogenesis is a multi-step process and begins with local degradation of the basement membrane. In order to grow, a tumour needs a vast blood supply for oxygen and nutrients. It is equally important for blood to supply oxygen and nutrients to both normal and tumour cells. Pro-angiogenic factors such as vascular endothelial growth factor (VEGF) and fibroblast growth factors (FGFs) are

upregulated during angiogenesis, while anti-angiogenic factors are down-regulated.

These signals can stimulate endothelial cells to construct capillaries within a tumour.

6. Ability to invade and metastasize Metastases are the cause of 90% of human cancer deaths. Metastasis is the spread of cancer cells to new areas of the body and is defined as the formation of secondary tumour foci. The classical simplification of metastasis steps includes: local invasion, intravasation, survival in the circulation, extravasation and colonization (Nguyen, Bos, & Massagué, 2009). In order to invade and metastasize to other parts of the body, the related gene expression level

involved in the regulation of cell-cell and cell-matrix interactions will be altered and cells undergo a process called epithelial-mesenchymal transition (EMT). For

example, loss of E-cadherin and acquisition of vimentin are two critical steps during EMT (Myong, 2012).

7. Ability to survive with hypoxia Even with angiogenesis, cells in the interior of a tumour may be in an oxygen-deprived situation called hypoxia. Hypoxia is

detrimental to normal cells as aerobic metabolism requires oxygen to convert glucose to energy. Cancer cells can switch from aerobic to anaerobic glucose

(27)

10

metabolism to allow cancer cells to produce energy and survive in oxygen-deprived conditions.

8. Escaping from the immune system The body’s immune system detects and destroys abnormal cells to protect the human body when they are functioning properly. Cancer cells are able to evade destruction by the body’s immune defenses, proliferate and eventually invade other tissues.

1.1.4 Breast cancer

Breast cancer is the most common cancer in women and 1 in 8 Canadian women is expected to develop breast cancer during their lifetime (Canadian Cancer Society, 2015). Breast cancer is not a single disease, but is comprised of many biological subtypes with distinct pathological features and clinical implications (Iwamoto & Pusztai, 2010; Tang, Wang, & Bourne, 2008). According to different histopathological and biological features demonstrated in breast cancer subtypes, the relevant therapeutic strategies may vary as well (Blows et al., 2010). Thus, it is clinically important to accurately group breast cancers into subtypes for therapeutic decision-making. Clinically, this heterogeneous disease is categorized into three therapeutic groups (The Cancer Genome Atlas Network, 2012): The estrogen receptor (ER) positive group is the most numerous and patients with ER⁺ status receive endocrine therapy. The human epidermal growth factor receptor 2 (HER2) Enriched (HER2E) group is a major improvement because of effective therapeutic targeting of HER2 with Herceptin. Triple-negative breast cancers (TNBCs, lacking

(28)

11

expression of ER, progesterone receptor (PR) and HER2) are a group with only chemotherapy as an option for treatment.

Gene expression profiling is a potentially powerful tool aimed at identifying the “molecular portrait” of invasive breast cancer, and breast tumours were

accordingly classified into four intrinsic subtypes with distinct clinical outcomes (Table 1.1) (The Cancer Genome Atlas Network, 2012). The rationale for this classification lies in the distinct gene expression patterns in each subtype and reflects the precise molecular level differences. Based on the recent development of high information content assays including DNA methylation, microRNA (miRNA) expression and protein expression, The Cancer Genome Atlas Network characterizes more completely the molecular architecture of breast cancer using six different technology platforms. The integrated molecular analyses of breast carcinomas significantly extend the knowledge base to produce a comprehensive catalogue of what is likely the genomic drivers of the most common invasive breast cancer subtypes (Table 1.1). The biological outcome of the four main breast cancer

subtypes caused by genetic and epigenetic abnormalities may indicate that plasticity and heterogeneity observed in clinic occurs within these major biological subtypes of breast cancer.

(29)

12

Table 1.1 Highlights of genomic, clinical and proteomic features of invasive breast tumour subtypes

(Modified from (The Cancer Genome Atlas Network, 2012) with permission).

Subtype Luminal A Luminal B Basal-like HER2E

ER⁺/HER2^-(%) 87 82 10 20

HER2⁺(%) 7 15 2 68

TNBCs(+) 2 1 80 9

TP53 pathway TP53 mut (12%);

gain of MDM2 (14%)

TP53 mut (32%);

gain of MDM2 (31%)

TP53 mut (84%);

gain of MDM2 (14%)

TP53 mut (75%);

gain of MDM2 (30%) PIK3CA/PTEN

pathway PIK3CA mut

(49%); PTEN mut/loss (13%);

INPP4B loss (9%)

PIK3CA mut (32%) PTEN mut/loss (24%) INPP4B loss (16%)

PIK3CA mut (7%); PTEN mut/loss (35%);

INPP4B loss (30%)

PIK3CA mut (42%); PTEN mut/loss (19%);

INPP4B loss (30%) RB1 pathway Cyclin D1 amp

(29%); CDK4 gain (14%); low expression of CDKN2C; high expression of RB1

Cyclin D1 amp (58%); CDK4 gain (25%)

RB1 mut/loss (20%); cyclin E1 amp (9%); high expression of CDKN2A; low expression of RB1

Cyclin D1 amp (38%); CDK4 gain (24%)

mRNA expression High ER cluster;

low proliferation Lower ER cluster;

high proliferation Basal signature;

high proliferation HER2 amplicon signature; high proliferation Copy number Most diploid;

many with quiet genomes; 1q, 8q, 8p11 gain; 8p, 16q loss; 11q13.3 amp (24%)

Most aneuploid;

many with focal amp; 1q, 8q, 8p11 gain; 8p, 16q loss;

11q13.3 amp (51%); 8p11.23 amp (28%)

Most aneuploid;

high genomic instability; 1q, 10p gain; 8p, 5q loss; MYC focal gain (40%)

Most aneuploid;

high genomic instability; 1q, 8q gain; 8p loss;

17q12 focal ERRB2 amp (71%) DNA mutations PIK3CA (49%);

TP53 (12%);

GATA3 (14%);

MAP3K1 (14%)

TP53 (32%);

PIK3CA (32%);

MAP3K1 (5%)

TP53 (84%);

PIK3CA (7%) TP53 (75%);

PIK3CA (42%);

PIK3R1 (8%) DNA methylation – Hypermethylated

phenotype for subset

Hypomethylated –

Protein

expression High oestrogen signalling; high MYB; RPPA reactive subtypes

Less oestrogen signalling; high FOXM1 and MYC;

RPPA reactive subtypes

High expression of DNA repair proteins, PTEN and INPP4B loss signature (pAKT)

High protein and phosphoprotein expression of EGFR and HER2 Percentages are based on 466 tumour. Amp, amplification; mut, mutation.

(30)

13

In summary, the hallmark characteristics distinguish cancer cells from normal ones. The transformation of a normal cell into a cancerous one requires deregulation of multiple cellular activities regulated by gene expression patterns.

Gene expression is mainly controlled by transcription regulation and post- translational regulation (see section 1.2).

1.2 Transcriptional regulation

Transcription is the process of RNA synthesis. The DNA code is transcribed into a sequence of messenger RNAs (mRNA), which are then translated to proteins.

Transcription factors (TFs) are sequence-specific DNA-binding factors involved in the process of transcription and are key cellular components that control gene expression. Thus, their activities determine how cells function and respond to the environment. Currently, there is keen interest in research into human

transcriptional regulation, but much remains to be explored.

1.2.1 Transcription factors recognize specific DNA sequences

Research in recent decades has contributed to the understanding of how TFs recognize their cognate binding sites in the genome and then initiate gene

regulatory functions. Structural analysis of protein-DNA recognition motif and sequence-dependent DNA recognition have revealed why many TFs preferentially bind to a specific DNA sequence (Rohs et al., 2010). The physical interactions between the amino acid side chains of the TFs and the accessible chemical and conformational signature of the base pairs determine the preference of TFs for a

(31)

14

given nucleotide at a typical position. These two recognition types are so-called base- and shape-readout, respectively. Furthermore, high-throughput datasets have revealed that TFs have distinct DNA-binding profiles, even when they exhibit a high degree of similarity in their DNA-binding domains. This means that they can

precisely regulate gene expression through temporal and spatial regulation (Noyes et al., 2008; Sibly et al., 2012).

The full picture of the assembly of multi-protein complexes on transcriptional regulation cannot be entirely provided by the high-throughput in vitro technology about specific individual TFs. Sequence-based computational models for describing the DNA-binding specificities of TFs are generated for predicting the binding

specificity to any new site (Zhou et al., 2015). These sequence-based DNA motif methods have the benefit of easily visualizing DNA sequence motif. However, these models only describe the DNA base readout by a TF and do not include the binding affinity. Recently, probabilistic models incorporating DNA structure-derived

features perform better than DNA-sequence based models (Sharon, Lubliner, & Segal, 2008). Hence, the integrated genomic and structural information about protein-DNA binding models is taken into both base- and shape-readout mechanisms.

1.2.2 Transcription factor regulation

The expression and or the activity of TFs themselves can be regulated. For example, the so-called “guardian of genome” p53’s gene expression is regulated by directing binding of several types of TFs (Saldaña-Meyer & Recillas-Targa, 2011).

(32)

15

Post-translational modifications (PTMs) is another way that can rapidly and reversibly regulate TF functions, including subcellular localization, stability and interactions with cofactors (Tootle & Rebay, 2005). Phosphorylation of E-twenty-six (ETS) family members, for example, at Serine/Threonine (S/T) residues in response to a variety of upstream signals, exerts broad spectrum effects on their activity. In addition, it has been shown that TFs activity may be regulated by glycosylation and some TFs are included in this cadre of targets including ETS transcription factor Elf- 1 is O-GlcNAc glycosylated (Juang, Tenbrock, Nambiar, Gourley, & Tsokos, 2002) and nuclear factor I (NFI) isoform which undergo N-glycosylation (Kane et al., 2002).

Other potential PTMs, such as acetylation and sumoylation are also involved in the activity regulation of TFs (Zhou et al., 2015).

Regulation of subcellular localization is another means to control the activities of TFs or other proteins with nuclear targets. Active nuclear import and export of transcriptional regulators are based on the recognition of specific signals in the protein sequence. A nuclear localization signal (NLS) in the protein sequence can direct it to the nucleus, while a nuclear export signal (NES) can lead the

molecule to be transported out of the nucleus (Nardozzi, Lott, & Cingolani, 2010).

Alternatively, subcellular localization can be regulated through a piggyback mechanism by binding to another molecule, transport signal masking, or by cytoplasmic retention (Cautain, Hill, Pedro, & Link, 2015). For example, activation- induced deaminase (AID), functioning as a mutator by deaminating cytosine and thus converting it into uracil, is unable to diffuse into the nucleus despite its small

(33)

16

size and its nuclear entry requires active import mediated by a conformational nuclear localization signal. In contrast, its C-terminus is a determinant for AID cytoplasmic retention (Patenaude et al., 2009). Another example is the cytoplasmic retention of hormone nuclear receptors. For instance, nuclear hormone receptor estrogen receptor α (ERα) is sequestered by metastasis-associated protein (MTA-1) in the cytoplasm and executes its non-genomic activity (Kumar et al., 2002).

Therefore, the subcellular distribution of TFs can determine their biological effect on the cell: genomic effects when they localize to the nucleus, and non-genomic effects in the cytoplasm.

1.2.3 Transcriptional repression and activation

The first and most fundamental order on gene regulation is achieved by the preferential binding of a TF to specific DNA sequence in promoters or enhancers.

Higher orders of regulation are accomplished by PTMs on TFs domains and recruiting chromatin-modifying enzymes to induce chromatin structural changes (Geertz & Maerkl, 2010). TFs bind to sequence-specific binding sites in the context of free DNA. However, when the recognition sites are buried in chromatin, TFs need to achieve proper binding by exploiting various strategies (Hahn, 2005) through their cofactors to regulate gene expression (Stampfel et al., 2015). Gene expression is not only regulated by TFs but also by epigenetic modifications.

Epigenetic modifications include DNA methylation and histone modifications.

DNA methylation is a post-replication modification predominantly found in

(34)

17

cytosines of the dinucleotide sequence CpG. Specifically, DNA methylation contributes to a silent chromatin state together with proteins that modify

nucleosomes (Jaenisch & Bird, 2003). The nucleosome is the fundamental unit of chromatin, and it is composed of an octamer of the four core histones (H3, H4, H2A and H2B). The histones N-terminal “tails” are unstructured and possess a large number of amino acid residues that are targets for PTMs, particularly lysine and arginine. There are at least eight distinct types of modifications found on histones, including acetylation, methylation (lysines, arginines), phosphorylation,

ubiquitylation, sumoylation and crotonylation. The extra complexity lies in that methylation at lysines or arginines could have three different forms: mono-, di-, or tri-methylation. Of all the known modifications, acetylation has the most potential to unfold chromatin. This vast array of epigenetic modifications allows an organism to respond to the environment through changes in gene expression.

In order to initiate transcription, nucleosomal DNA has to disassemble first.

The cooperative TFs binding, chromatin-remodeling complexes and actively

transcribing Pol II can all mediate histone displacement. The cofactors of TFs can act as activators or repressors on gene regulation. For example, histone

acetyltransferases (HATs) and histone deacetylases (HDACs) are two counteracting enzyme families controlling the acetylation state of the lysine residues of the core histones. The acetylation of lysine residue removes the positive charge on the histones and thereby the interaction of the N termini of histones with the negatively charged DNA decreases. As a consequence of acetylation, the condensed chromatin

(35)

18

is transformed into a relaxed structure which facilitates gene transcription. But this relaxed state can be reversed by deacetylation, which is performed by HDACs. Once the lysine is deacetylated, the chromatin is back to the condensed state. Acetylation or deacetylation cannot be done by HATs or HDACs alone and is always coupled by other molecules to complete this reaction. For example, the deacetylase activity of HDAC3 strictly requires interaction with its transcriptional co-repressor nuclear receptor co-repressor (N-CoR) (Zhou et al., 2015).

1.2.4 Roles of transcription regulators in cancer

Many TFs are inactive under normal physiological conditions and their expression and activity are tightly regulated. A high proportion of oncogenes and tumour suppressor genes encode TFs. Many human cancers are dependent on the inappropriate expression or activation and inactivation of TFs as well as mutation.

For instance, somatic mutations in the p53 gene are some of the most frequent alterations in human cancers (Olivier, Hollstein, & Hainaut, 2010). Hence, TFs represent highly desirable and logical points of therapeutic interference.

Cellular signal transduction induced by the genetic and epigenetic changes is dysregulated in cancer cells. In each pathway, the extracellular signal is received by a receptor and conveyed into the nucleus. TFs and their cofactors are at the end of the signalling pathway, which can regulate gene expression or repression (Nebert, 2002). In clinical trials, progression-free survival of patients with cancers who were previously regarded as untreatable, were improved by drugs that target intracellular

(36)

19

signalling pathways. However, alternate signalling pathways that are not targeted by drugs or a downstream mutation within the kinase-mediated signalling cascades has curtailed the benefit (Gonda & Ramsay, 2015). The cancer phenotype is not only defined by the misregulation of key transcriptional regulators but the misregulation is also critical for cancer development and maintenance. It can therefore be

proposed that when these transcriptional regulators act as therapeutic targets, they are less prone to be bypassed by an alternative pathway (Gonda & Ramsay, 2015).

Some transcriptional regulators are already under-investigation as potential therapeutic targets (Berg et al., 2002; Yardley et al., 2013). For example, a vector- based DNA Myb vaccine showed some antitumour efficiency against the metastatic spread in a model of mammary cancer (Carpinteri, 2012).

Multicellular organisms are strictly ordered and require extensive coordination and communication between cells; many TFs and cofactors are

involved in this hierarchical communication. In response to external stimuli, TFs and cofactors turn on/off appropriate gene expression. Growth factors act as one of the external stimuli which can activate multiple pathways and lead to pleiotropic effect on cell biology.

1.3 Growth factors and their function

Cellular phenomena—proliferation, differentiation, migration and survival/death are not autonomous; much of this is regulated by extracellular proteins (growth factors) that positively and negatively regulate these actions.

(37)

20

Regulation is achieved via transmembrane receptors that growth factors bind on the extracellular surface of cells in order to transduce cellular signalling events in the cytoplasmic compartment (Lemmon & Schlessinger, 2011). Once the ligand binds to the respective receptors, it will trigger intracellular signalling events in both

transcription-dependent and independent pathways in the target cells (Brunet, Datta, & Greenberg, 2001). Briefly, growth factors are divided into cytokines, and polypeptide growth factors (Vlasova & Bohjanen, 2016), both of which affect nearly every biological process (Vlasova & Bohjanen, 2016). Cytokines, often compared with growth factors, are a class of signalling molecules that primarily affect the cells of the immune system (An, 2009). From here on, we will mainly focus on

polypeptide growth factors that affect most cells of the body.

1.3.1 Growth factors and their receptors

Polypeptide growth factors types: Polypeptide growth factors can act by multiple means paracrine, endocrine and autocrine systems (Hull & Harvey, 2014). There are multiple “superfamilies” of peptide growth factors that contain subfamilies of

proteins, with related primary sequences. For example, fibroblast growth factor (FGF) superfamily contains at least 22 distinct members (Zhang et al., 2006).

Growth factors are ligands for transmembrane receptors. Each growth factor superfamily has a corresponding family of related receptors with high specificity.

Family members can bind to a single receptor and there are also ones that bind to multiple receptors. For instance, the aforementioned FGF family of 22 structurally-

(38)

21

related molecules can bind to four high affinity, ligand-dependent FGF receptor tyrosine kinase molecules (FGFR1-4) (Zhang et al., 2006). The activation of FGFR results in the stimulation of various signal transduction cascades implicated in multiple aspects of embryonic development, tumour growth, angiogenesis, wound healing, and physiology (Ornitz & Itoh, 2001; Powers, McLeskey, & Wellstein, 2000).

Growth factor receptors: Growth factor receptors are plasma membrane-spanning proteins that bind with a specific growth factor on the external surface of a cell and transduce a signal that regulates cell division. They contain an intracellular domain with enzymatic function that is activated by growth factor binding. For example, epidermal growth factor (EGF) is an approximately 6 kDa molecule and binds to a 170 kDa plasma membrane receptor (EGFR) resulting in receptor dimerization, autophosphorylation (in trans) and activation of various downstream signalling pathways (Zhang et al., 2006). Growth factor receptors also define cancer

hierarchies (Venere, Lathia, & Rich, 2013) and increased expression or activation of receptor tyrosine kinases occur frequently in human breast carcinomas. For

example, breast cancers are classified into different subtypes depending on the expression of ER, PR or Her2 (refer to Table 1.1). Epithelial breast cancer cells are well recognized as commonly over-expressing the Insulin-like Growth Factor-1 (IGF- I) receptor (Christopoulos, Msaouel, & Koutsilieris, 2015), which is a high-affinity receptor for both insulin and IGF-I (Belfiore & Frasca, 2008). EGFR is frequently over-expressed in TNBC (Nakai, Hung, & Yamaguchi, 2016) and an over-expression

(39)

22

of EGFR is correlated with poor prognosis of colon cancers as well (Sasaki, Hiroki, &

Yamashita, 2013).

1.3.2 Growth factor function

A. Growth factors can regulate proliferation: Few cells can proliferate without the stimulus of growth factors. Thus, in this regard, growth factors play a significant role during development. The presence of IGF-1 dramatically enhanced early stage proliferation of EGF/FGF-responsive neural stem cells in vitro (Supeno et al., 2013).

In other cases, growth factors (i.e. transforming growth factor β (TGFβ)) can inhibit cell proliferation (Zermati et al., 2000).

B. Positive and negative regulation in development by GFs: The complexity of embryogenesis is reflected in the presence of complex interactions between growth factor signalling pathways. Recent studies have demonstrated that growth factor receptors are expressed by pre-implantation embryos and growth factor deprivation can result in suboptimal growth as well as developmental abnormalities (Richter, 2008). Pre-implantation embryos also express many growth factors of their own, including EGF, insulin-like growth factor1 (IGF-1), IGF-2, VEGF, platelet-derived growth factor (PDGF) and fibronectin (Richter, 2008). These autocrine growth factors are believed to be the primary reason for embryonic development (Richter, 2008). In some cases, growth factors and their receptors can act as development inhibitors. An example of that is Met, the receptor of Hepatocyte growth factor (HGF), which regulates skeletal muscle differentiation; a novel spliced isoform

(40)

23

Δ13Met has been found, which can inhibit muscle cell differentiation (Park et al., 2015).

C. Growth factors regulate transcription: Many different types of stimuli can cause the activation of protein kinases, which can further affect gene expression. Growth factor-dependent MAPK and PI3K/AKT pathway activation can phosphorylate many downstream transcription factors. These phosphorylated transcription factors (TFs) are activated and will further up-regulate or down-regulate gene expression

(including other transcriptional factors’ gene expression).

D. Growth factors and wound healing: Wound sites release several growth factors, including IGF, EGF, FGF, PDGF, TGF and so on. The clinical application of growth factors to stimulate the healing of wounds is currently being investigated (Grazul- Bilska et al., 2003).

Taken together, a cell receives extracellular signals through ligand-receptor interaction and the signal is sensed through activation of related pathways in the cytosol. The signals then have to be transduced into the nucleus, where signals eventually reflect on the genomic level of gene expression/repression. The signal, from the external to cytosol to nucleus, is through tiers of activation and

translocation of molecules.

(41)

24

1.4 Nuclear-cytoplasm exchange system

A cell consists of different cellular compartments, which are associated with a diverse range of biochemical processes (Sprenger et al., 2008). Protein function is always related to its subcellular localization and proteins must be targeted to the appropriate compartment to ensure proper function. Therefore, a protein’s cellular role may be inferred by localizing to distinct compartments (Rezácová, Borek, Moy, Joachimiak, & Otwinowski, 2008). Understanding protein subcellular localization is not only important for elucidating its function in cells, but also for the organization of the cell as a whole (Scott, Calafell, Thomas, & Hallett, 2005).

The nuclear membrane separates nuclear and cytoplasmic compartments in eukaryotic cells and supports as a structural frame of the nucleus. The nuclear membrane, acting as a barrier between cytosol and nucleus, prevents the free nuclear to cytoplasmic diffusion of molecules; such movement is directed by signals to translocate (Cooper & Hausman, 2000). Nuclear pore complexes (NPC) penetrate through the nuclear membrane and serve as a transporting channel for

macromolecules between the two compartments. Many essential regulatory molecules are shuttled between compartments; as example, histones and TFs are imported into the nuclear compartment, while transfer RNA (tRNA), ribosomal RNA (rRNA), and messenger RNAs (mRNA) are transcribed in the nucleus and exported out to the cytoplasm where they function in translation (Beck & Hurt, 2016).

(42)

25 1.4.1 Nuclear pore complexes

Macromolecules can transiently dock and interact with nucleoporins during the transportation process. NPCs create an aqueous channel through which

macromolecules are transported (Adam, 2001). The NPC is a large protein complex that can be easily detected by electron microscopy and acts as a molecular

transportion gate between cytoplasm and nucleus (Appen & Beck, 2016). The mass of NPCs in higher eukaryotes is about 125 MDa and the proteins that comprise the complex are called nucleoporins (Nups) (Adam, 2001). Each NPC is composed of about 50-100 different Nups. Morphologically, NPCs contain a membrane-embedded central core structure, cytoplasmic and nuclear extensions which form cytoplasmic filaments and nuclear baskets, respectively, which act as cargo docking sites (Beck &

Hurt, 2016) (Fig. 1.2). The membrane-embedded central core contains three stacked rings. The middle ring spans and crosses the fused inner and outer nuclear

membranes and is sandwiched by the cytoplasmic and nucleoplasmic rings from both distal ends (Beck & Hurt, 2016). The cytoplasmic ring constitutes of eight 50 nm filaments and the nuclear ring is connected to a basket-like assembly of eight thin terminal rings (Fig. 1.2). Macromolecules bearing transport signals translocate through the center of an NPC gate.

The number of NPCs depends on the demands of cells for nucleocytoplasmic exchange and varies dramatically with cell size and the demands of cellular activities, such as proliferation. For example, there are about 3000-5000 NPCs in a

proliferating human cell (Kabachinski & Schwartz, 2015). The commonly used

(43)

26

human cell line, Hela, contain on average 3000 NPCs in each nucleus (Kabachinski &

Schwartz, 2015).

As the sole gateway for the exchange of material between nucleus and cytoplasm, NPCs support two modes of transport: passive diffusion and receptor- facilitated translocation (Naim et al., 2007). Small molecules such as metabolites can passively diffuse through the pore without assistance, but it becomes increasingly slow as the size of the particle approaches ~10 nm in diameter, which corresponds to a protein with a molecular weight of about 45 kDa (Naim et al., 2007). Passive diffusion is only reasonably fast for proteins smaller than 20-30 kDa. In contrast, larger proteins, RNAs, and their complexes require active transport into the nuclues.

But not all molecules whose molecule weight is less than 20-30 kDa will diffuse passively in the cells. For example, histones and tRNAs enter the nucleus through carrier-mediated transport, even though their molecular weight is less than 20-30 kDa (Suntharalingam & Wente, 2003).

(44)

27 Figure 1.2 Schematic representation of the NPC

(Modified from (Grünwald, Singer, & Rout, 2011) with permission) Cytoplasmic and nuclear extensions of the NPC's periphery are indicated on the cytoplasmic and the nuclear surface.

(45)

28

(46)

29 1.4.2 Transport machinery

Facilitated transport requires specific interactions between the molecule being translocated and the NPC. This nucleocytoplasmic transport is mainly mediated by transport receptors belonging to the superfamily of importin-β-like proteins called karyopherins. Based on the direction to which these receptors carry their cargo, they are classified as importins or exportins which can directly interact with the proteins on the surface of NPCs. As the name implies, importins are

accountable for directing the cargo to the nucleus, whereas exportins shuttle the molecules from nucleus to cytoplasm (Yuh & Blobel, 2001).

Nucleoporins (Nups) are often grouped into three types: (i) transmembrane Nups which anchor the NPCs in the nuclear envelope; (ii) phenylalanine-glycine repeats (FG-repeats), and (iii) structural Nups, which act as a scaffold to interact with transmembrane Nups and FG-Nups. Amongst them, FG-Nups play dual roles in nucleocytoplasmic exchange: first, they function as a permeability barrier of the NPCs; second, karyopherins can transiently interact with FG-Nups and transport the cargo through NPCs, which the FG-Nups support as an anchor (Wälde & Kehlenbach, 2010).

Another part of the machinery is Ran(Ras-related nuclear protein), a member of the Ras family of small G proteins, which is essential for the translocation of proteins and RNA through NPCs (Sazer & Dasso, 2000). The prime function of Ran is

(47)

30

to regulate the binding of cargo molecules and will be further discussed in section 1.4.3.2.

1.4.3 Cargoes and signals

Eukaryotic cells must accomplish the rapid and receptor-mediated transport of thousands of proteins and RNAs into and out of the nucleus and karyopherins are taking care of this bidirectional transport (Pemberton & Paschal, 2005). This parcel- like delivery model raises the question of how karyopherin: cargo recognition occurs. Many studies have shown proteins that undergo nuclear import or export generally contain a NLS or NES, respectively (Lange et al., 2007; Weis, 2003).

1.4.3.1 NLS-dependent and independent nuclear import

In 1984, a NLS was first characterized from SV40 Large T antigen and consisted of a short sequence of basic amino acids (Dingwall & Laskey, 1991). The NLS region in SV40 Large T antigen had a stretch of five basic amino acids

127PKKKRKV¹³³ and was defined as monopartite NLS (Kalderon, Richardson, Markham, & Smith, 1984). Subsequently, a related signal of two basic clusters separated by about ten residues was identified in Xenopus nucleoplasmin and defined as bipartite (Dingwall, Robbins, Dilworth, Roberts, & Richardson, 1988).

The sequences identified in SV40 T-Ag and Xenopus nucleoplasmin are now referred to as classical NLSs (cNLSs) and require the karyopherins importin-α and importin-β for nuclear transport (Lange et al., 2007).

(48)

31

Consecutive residues from the N-terminal lysine of monopartite NLS are referred to as P1, P2 and a lysine in P1 position is mandatory for monopartite cNLS (Conti & Kuriyan, 2000; Fontes, Teh, & Kobe, 2000; Hodel, Corbett, & Hodel, 2001), followed by basic residues in positions P2 and P4 to yield a consensus sequence of K-K/R-X-K/R, where X stands for any amino acid. Quantitative analyses of the accumulation percentage of nuclear import of eGFP fused with different NLSs demonstrated that a monopartite NLS is more efficient than a bipartite NLS (Ray, Tang, Jiang, & Rotello, 2015). Since the discovery of the NLSs in SV40 T antigen and nucleoplasmin, many other NLSs have been described, as well.

The formation of the import complex is mediated by specific sites on importin-α by recognizing the NLS in the molecules to be imported (Conti, Uy, Leighton, Blobel, & Kuriyan, 1998). Importin-α is composed of a tandem series of Armadillo (ARM) repeats which form a banana-like molecule, producing a curving structure with two NLS-binding sites (Stewart, 2007). Classically, the adaptor protein importin-α recognizes the NLS present in the cargo and forms a dimer with importin-β. Importin-α binds to importin-β through a domain known as the

importin-β binding (IBB) domain, which is located in the N-terminus. IBB can compete with NLS and replace the NLS-binding sites, leading to the release of cargo proteins (Lott & Cingolani, 2011). The cargo:importin-α:importin-β complex is transported through NPC and dissociated by the binding of RanGTPase.

The import of many nuclear proteins is thought to be mediated by the classical NLS. However, it is now accepted that import signals unrelated to the

(49)

32

classical NLS exist (Freitas & Cunha, 2009). These NLS-independent molecules could be imported through interaction with other proteins that contain a functional NLS, called a piggyback mechanism, or rely on importin-β-related molecules. For instance, β-catenin, which mediates a late step of the Wnt/Wingless pathway, is imported into the nucleus by binding directly to importin-β or β-like import factors (Fagotto, Gluck,

& Gumbiner, 1998).

1.4.3.2 NES-dependent and independent nuclear export

Typically, basic residues (e.g. K, R) are enriched in NLSs. In contrast, a leucine-rich nuclear export signal (NES) is present in cargos exported to the

cytoplasm. NES also contain critical hydrophobic residues, which are necessary for recognition by the nuclear export receptor CRM1 (Fung, Fu, Brautigam, & Chook, 2015). The most conserved NES pattern is the LXXXLXXLXL motif where “L” is a hydrophobic residue (Leucine) and “X” is any other amino acid. The spacing

between the hydrophobic amino acid residues varies, although the most conserved pattern is LXXLXL, while some fit the LXXXLXL pattern. However, it has been

determined that approximately 15% of protein NESs do not conform to either of the LXXLXL or LXXXLXL patterns, indicating a significant degree of flexibility in the export signal (L. Cour et al., 2004).

CRM1 is an essential exportin utilized in all types of cells, and it exports numerous cargos including proteins and RNAs (Cullen, 2003; Delaleau & Borden, 2015). NESs bind weakly to CRM1, which ensures that once transport is completed,

(50)

33

these cargos are easily released (Fischer et al., 2015). The way molecules bind to CRM1 through NESs for exportation is called the canonical pathway; also, a non- canonical CRM1 export mechanism exists as well.

During the whole cycle of nucleocytoplasmic shuttling, Ran plays a major role in assisting loading or discharging cargo. For nuclear import, the cargo molecule with an accessible NLS binds to an importin molecule and this complex transports into the nucleus through NPC (Cingolani, Petosa, Weis, & Müller, 1999; Conti et al., 1998). RanGTP can then bind to importin and cause the dissociation of imported complexes by direct or indirect competition (Chook & Blobel, 1999; Cautain et al., 2015). Subsequently, the RanGTP-importin complex is recycled to the cytoplasm.

Conversely, RanGTP binds to CRM1 and promotes the tight assembly of exported complexes in the nucleus (Cassar et al., 2007). Once the complex is exported to the cytoplasm, RanGTP is hydrolyzed to RanGDP by RanGAP. RanGDP weakens the affinity between NES and exportin, causing the dissociation of cargoes (Koyama &

Matsuura, 2010). RanGDP is then recycled in the cytoplasm by Nuclear Transport Factor 2 (NTF2) back to the nucleus where Ran is loaded with GTP by the guanine nucleotide-exchange factor regulator of chromosome condensation (RCC1). The import and export cycle is illustrated in Fig. 1.3.

(51)

34 Figure 1.3 Ran directs nucleocytoplasmic transport

(Modified from (Clarke & Zhang, 2008) with permission) (A) The GTP–GDP cycle of Ran. Ran is loaded with GTP by the guanine nucleotide-exchange factor RCC1. RanGTP adopts a distinct conformation that allows it to interact with a transport factor from the importin- β superfamily, also known as the karyopherins. Hydrolysis of GTP to GDP by Ran requires the interaction of a Ran GTPase-activating protein, RanGAP1, and is stimulated by Ran- binding protein-1 (RanBP1) or RanBP2. RanGDP has a different conformation that does not interact strongly with karyopherin and can be considered inactive. Mutants of Ran block the GTP–GDP cycle: RanT24N has a reduced affinity for nucleotides and forms a stable complex with RCC1, thereby blocking RanGTP formation, whereas RanQ69L cannot hydrolyse GTP and is locked in the GTP-bound conformation. (B) Ran shuttles across the nuclear envelope through nuclear pores, but is concentrated in the nucleus because of nuclear transport factor-2 (NTF2)-mediated active import. In the nucleus, a high concentration of RanGTP is generated by nucleotide exchange. This is catalysed by chromatin-bound RCC1 and might be promoted by the nucleotide dissociation factor MOG1 and the accessory factor RanBP3 (not shown). RanGTP causes the dissociation of imported complexes, which contain proteins that carry a nuclear localization signal (NLS), by binding to importin-β and ejecting the cargo.

Conversely, binding of RanGTP to chromosome-region maintenance protein-1 (CRM1) promotes the assembly of export complexes containing proteins with a nuclear export signal (NES). In the cytoplasm, RanGTP meets RanGAP1 and RanBP1 or RanBP2, which stimulates GTP hydrolysis and the export complexes dissociate. The importins and exportins are recycled by transport back across the pore.

(52)

35

REGULATION OF NUCLEOCYTOPLASMIC SHUTTLING OF THE TRANSCRIPTIONAL REGULATOR HUMAN MESODERM INDUCTION EARLY RESPONSE 1 α IN BREAST CARCINOMA CELLS